Guidelines for Sequencing PFM Reforms

shrewdnessfreedomSoftware and s/w Development

Dec 2, 2013 (3 years and 10 months ago)

236 views


Background Paper 1



Guidelines for Sequencing PFM Reforms



Jack Diamon
d

Ju
ly

2012









THIRD

DRAFT
*

Prepared as a backg
round paper to The Guidance

Note in Sequencing PFM Reforms, and
should be read in conjunction with

Background Paper 2,

Core
PFM Functions and PEFA
Performance Indic
ators
,


by Daniel Tommasi, June

2012

* Revised in light of comments received following PEFA Steering Committee June
2012 and
FAD course on sequencing presented at the Singapore Institute, July 2012




2


Contents

Page



PART A. The Approach to PFM Reform Sequencing
................................
................................
.............

4


I: Sequencing PFM Reforms: The State of the Art

................................
................................
......................

4

Does Sequencing in PFM Reforms Have any Practical Meaning?


................................
.................

4

A Review of the PFM Sequencing Literature

................................
................................
..................

5

What Does The Literature Tell Us?

................................
................................
...............................

11

II: A Framework for Interpreting the Sequencing Literature

................................
................................
.....

16

A Suggested Framework for Reviewing the Sequencing Literatur
e

................................
..............

16

The Hierarchy among Top Level PFM Priorities

................................
................................
..........

20

Using the Framework to Re
-
Interpret the Sequencing Literature

................................
..................

23

Areas to Be Addressed in this Background Paper

................................
................................
.........

25

PART B. An Overview of Sequence in the Development of a PFM System

................................
........

28

III: Step One: Establishing “Core” PFM Functions as the First Priority

................................
...................

28

The Need to Move Away From “Basics” as a Reform Target

................................
.......................

28

Defining What Constitute “Core” PFM Processes

................................
................................
........

30

Using the PEFA Analytical Framework to Measure Core Functions

................................
............

32

PEFA as a Tool for Determining Reform Sequence

Has Limitations

................................
...........

37


IV:
Step Two: Str
engthening the IT, Accounting,
and the
Legislative Base for PFM Reforms

................

40

Establishing a Solid Platform for Moving “Beyond The Core”

................................
....................

40

The IT Basis for PFM Reforms

................................
................................
................................
.....

42

The Accounting and Reporting Requirements in PFM Reform
................................
.....................

49

The Legal Framework for PFM Reform

................................
................................
........................

57

V: Step Three: Move from Annual to Medium
-
Term Budget Planning

................................
....................

61

Introduction: The Need for Medium
-
Term Planning

................................
................................
.....

61

Preconditions for Moving from Annual to Medium
-
Term Budget Planning
................................
.

63

Stylized Development Path for Medium
-
Term Budget Planning

................................
..................

66

Some
C
larifications and
Q
ualifications

................................
................................
.........................

71

VI: Step Four: Program and Performance Budgeting

................................
................................
................

74

The
C
ritical
R
ole of
P
rogram
-
B
based PFM

................................
................................
...................

74

Phased Introduction of Program Budgeting

................................
................................
...................

77

Final Phase: Moving to a Performance Based Budget System
.

................................
.....................

83

PART C. Deciding the Sequencing of PFM Reform Actions

................................
................................

91

VII: Sequencing as One Component in Reform Design

................................
................................
............

91

A Schematic View of The Reform Decision Process

................................
................................
....

91

Stage One: Analysis and Diagnosis

PFM Technical Factors
.

................................
.....................

92

Stage Two: Analysis and Diagnosis


Factors External to PFM.

................................
.................

93

Stage
T
hree: Settle on a sub
-
set of
reform options.

................................
................................
.......

95

Stage
F
our: In
D
ialogue with interested parties come to
A
greement on
R
eform
A
ctions.

...........

95

Stage
F
ive: Decide
S
equencing of
R
eform
A
ctions.

................................
................................
......

96

Stage
S
ix: Put in place an
A
dequate
D
elivery
M
echanism.

................................
...........................

98

Stage seven: Formal
M
onitoring and
R
eview of
R
eform
E
fforts.

................................
.................

99

Some Messages Derived from the Literature
.

................................
................................
..............

100

VIII: Assessing Reform Risks at the Top Political Economy Level

................................
.........................

101

Change Management is Essential to Address the External Factors

................................
.............

101

Deciding on What is
Possible in PFM Reform

................................
................................
............

103

Change Management as Risk Management

................................
................................
.................

106

Reform Implications in Deciding the Feasibility of Reforms

................................
......................

109

IX: Assessing Reform Risks at the Institutional and Organizational Levels

................................
...........

118

3

An Approach to Assessing Risks From External Facto
rs at the Lower Level

.............................

118

The Institutional Structure of PFM Systems

................................
................................
................

119

The Internal Organization of PFM Systems

................................
................................
................

127

X: A Risk
-
based Approach to Sequencing PFM Reforms

................................
................................
........

134

Reform Risk as Criterion for Sequencing Decisions

................................
................................
...

134

Making the Risk Based Approach Operational

................................
................................
............

136

Employing the Risk Analysis to Determine Reform Strategy

................................
.....................

138

Choosing the Sequence of Reform Actions

................................
................................
.................

140

Concluding
R
emarks

................................
................................
................................
..............................

145

Annex: Sequencing Program/Performance Budgeting Reforms

................................
..........................

87

Tables

8.1 Inventory of Political Economy Influences on PFM Reforms

................................
............................

111

9.1. Factors Arising from Structure of the PFM System
................................
................................
...........

126

9.2
.
A. Top Management Profile

................................
................................
................................
..............

128

9.2.

B Middle management
profile

................................
................................
................................
...........

128

9.2
.

C. Operating level profile

................................
................................
................................
..................

129

9.3. Some Key Factors

Affecting Reform Arising from Internal Organization of PF
M Entries

..............

13
5


10
.
1

Examples of
R
isk

L
evels from Different types of PFM Reform in same External Environment


....

137

10.
2
. Risk Impact Associated with Different Categories of External Factors and Type of Reform

.........

138

Figures

2.
1
.
Schematic Overview of the PFM Conceptual Framework based on Three Deliverables

....................

23

2.2
.

Reform Priorities

................................
................................
................................
................................
..

21

3.1. Aligning Core Functions with PFM Deliverables

................................
................................
................

32

4.1. Three
-
Tracked Sequencing of PFM Reforms Beyond the Core

................................
..........................

41

6.1. A Strategy to Move to Perfor
mance budgeting
................................
................................
...................

.87

7.1. Schematic Decision Framework

................................
................................
................................
..........

91

7.2. Classification of Political Economy Factors in Reform

................................
................................
.......

94

8.1. Deciding on Available Reform Options

................................
................................
.............................

103

9.1. Reform Space

the Three

"A"s

................................
................................
................................
.........

119

10.1. Risk Tolerability Matrix

................................
................................
................................
..................

135

10.
2
. The Trade
-
off Between Reform Impact and Risk

................................
................................
............

140

10.
3
. The Risk Based Approach to Sequencing

Reforms

................................
................................
.........

143

Boxes

1.1. The Basics First Approach

................................
................................
................................
................................
.............

6

3.1. PFM Functions are Required to Ensure a

Core Level of Compliance

................................
...........................

31

3.2. PEFA Indicators Allo
cated to PFM Stages and Target Scores for LICs

................................
.......................

34

3.3. PEFA Indicator Scores for Different Classes of Countries

................................
................................
...............

36

3.4. Core Functions and Their Target PEFA Scores

................................
................................
................................
....

36

4.1 The Use of a FMIS in PFM Involves more than Just Automation

................................
........................

45

4.2. Sequencing the Design and Implementation of FMIS Projects

................................
................................
........

48

4.3. Steps in Moving to Modified Accrual

................................
................................
................................
......................

52

4.4. Required Provisions in the Legal Framework to Meet

................................
................................
.......................

5
7

5.1. Phased Development of Medium
-
Term Budget Planning

................................
................................
.................

65

8.1. PFM Strategies in Post
-
Crisis and
Fragile States
................................
................................
...............................

110

References


................................
................................
................................
................................
........

147



4

PART A.

THE APPROACH TO PFM
REFORM SEQUENCING



CHAPTER I:
SEQUENCING PFM
R
EFORMS: THE
S
TATE OF THE
A
RT


Summary:

This chapter reviews recent key contributions to the PFM sequencing
literature, and
critically examines some of the dominant views found there. In light of this review, it highlights
common themes that emerge in the literature.


DOES
SEQUENCING IN
PFM
REFORMS HAVE ANY PRA
CTICAL MEANING?

Despite some recent attempts to agre
e a sequence in planning PFM reforms, a consensus
remains lacking.
Indeed, some would deny the concept has any practical use.

At first glance,
both historically and across countries, the argument for a technical solution to sequencing PFM
reforms seems weak. It has been pointed out that if we look historically at how advanced
countries have advanced their PFM systems progress app
ears haphazard, spasmodic,
opportunistic, and politically motivated. Certainly their reforms hardly could be characterized as
technical exercises that would merit being called a planned sequence of events. Further, the vast
differences in country context s
eem to suggest a common reform approach, or objective, is likely
to be difficult to identify and has given rise to further doubts. How can PFM reform be
generalized to agree a sequence if non
-
technical factors dominate? After all no country is exactly
like

another now, so why should it be in the future? Can all countries be viewed as being on the
same reform path? Should reform sequencing be viewed as a process to ultimately attain current
best, or even good practice, as defined by advanced countries, when
experience shows that best
practice is constantly changing? More fundamentally, does

best practice


have much relevance
to the majority of countries?

A
lthough different advanced countries have taken different reform routes the end result
has not been too
different.

As a consequence the concept of

good


practice in PFM seems, if
not in the details, to be broadly agreed. On this agreement international assessment tools, like
PEFA, have been founded
.
Admittedly, while the advanced countries have reached simi
lar end
results, they have displayed a variety of paths and completely different time
-
lines in reaching
these results
.
But this great variation does not mean that the experience of these pioneers is not
relevant, and that others undertaking PFM reforms hav
e nothing to learn from it.

T
o be successful in PFM reform there is little choice but to attempt to define a desirable
sequence of reforms.

It is admitted this paper takes a decidedly positive stance to the
sequencing issue .
If we cannot define at least a
notional reform path, no reform can define what
it is ultimately attempting to achieve, and certainly cannot define the steps to be taken on the way
to achieving this. In the absence of this

vision


reform efforts, and the associated donor support,
will t
end to be band
-
aid, quick
-
fix, dealing with symptoms, and reforms will be difficult to
5

sustain. Apart from the need to define the broad objectives of reform, the practical problems in
implementing PFM reforms also suggest it is not possible to avoid the se
quencing problem. PFM
is a system so wide in nature that not all elements can be tackled at once. This implies that, of
necessity, reformers need to prioritize to pursue a viable reform path. How to prioritize lies at the
heart of the sequencing issue.

A
REVIEW OF THE PFM SE
QUENCING LITERATURE

This takes

the

basics first


approach as a starting point
.
In many ways it is difficult to
differentiate the literature on PFM sequencing from the more general literature on the design of
PFM reform programs. However, there are a few important recent works which directly address
the issue of the order in which ref
orm activities should be undertaken. In this subset, Allan
Schick’s1998 contribution, stressing the important principle of

getting the basics right


as a first
priority when undertaking reform, has already gained wide currency. His was originally a plea
f
or reversing what he felt were the over
-
ambitious attempts to establish PFM international best
practices in less developed countries which lacked the capacity to operate even basic processes.
His message was summed up succinctly in the World Bank PEM Handb
ook (see Box 1.1).

However, the

basics first


approach has not proved popular.

Notwithstanding its broad
endorsement by the majority of donor agencies
1
, this very simple message on sequencing, that
low capacity countries should first focus on their basic
PFM processes, has been surprisingly
controversial for a number of reasons. First, it has been difficult for experts in the field to agree
on what should be considered

basic

, although there is a substantial degree of overlap in all
interpretations.
2

Seco
ndly, others have objected to designating any PFM process as

basic


as
involving a large degree of arbitrariness, and hence the approach gives insufficient guidance to
reform action.
3

Thirdly, that the concept of sticking to the

basics first


principle i
s not an easy
strategy to sell within LICs, as well as within the donor community.
4







1

Among them, DFID (2001,pp.46ff.), USAID
(Browne, 2010,Chapter 5), ADB (Schiavo
-
Campo,1999,p.118ff),
IBRD (
1998).

2

Compare, for example, the

basic


reforms identified by Tandberg and Pavesic
-
Skerlap, 2008, with Schick’s
checklist. See also, Tommasi, 2009, p.22, for his definition of

basics


ve
rsus

beyond basics

, and Browne, 2010,
pp.16ff, for a description of her interpretation of

first level


or basic reforms.

3

Allen, for example, sees the concept of

basic


as subjective, as he puts it

it is similarly unclear what criteria
should be

applied in determining whether a country has achieved sufficient progress in improving its basic systems
to move forward to more advanced measures. Should a government be encouraged to start work on such reforms
before
all
the basics are in place? Are som
e of the

basics more important than others in establishing

essential
preconditions for moving on to the menu of more advanced measures?


(Allen 2009,p.17)
.

4

Allen and others have noted how unattractive the

return to basics


strategy can be for politician
s knowledgeable
about the latest reforms in advanced coun
tries and eager to show results

(Allen, 2009; Andrews 2006).

6



Alongside the

basics fir
st


approach, and perhaps recent
ly the approach
that has gained
wide interest, is the

platform approach
.


A
ssociated with Peter Brooke (2003)
5
, t
his strategy
for sequencing PFM reforms proposes

to re
place of the past emphasis
on individual reform
activities. It argues

that a more productive approach is to focus on reform activities packaged
together. These packages of supportive measures then should form a logical sequence for reform,
so that on
ce one package of activities is completed it will form the basis, or

platform
,

on which
to anchor

a further package of complementary reforms. This strategy is designed to advance the
PFM system over a period, perhaps as long as ten years, with each
platform’s activities lasting
for a period of two to three years. The total strategy is envisaged as being up
-
dated and rolled
over, say, every two years depending on progress made. There have been a number of arguments
advanced in support of such an appro
ach to sequencing reform. First, is the simply technical
one

that some PFM processes are technically dependent on others being in place to ensure



5

A variant of the platform approach is the so
-
called

evolutionary approach


developed for Ethiopia and
implemented by Harvard University

(see Petersen,

2007). Although differing in content, this approach is little
different in design than that advanced by Brooke.

Box 1
.1.

The

Basics First


Approach

In elaborating his argument for

䝥瑴dng 瑨攠B慳楣s oigh琬


卣p楣k s瑡瑥sW



The
government should foster an environment that supports and demands performance before
introducing performance or outcome budgeting.



Control inputs before seeking to control outputs.



Account for cash before accounting for accruals.



Establish external control
s before introducing internal control.



Establish internal control before introducing managerial accountability.



Operate a reliable accounting system before installing an integrated financial management system.



Budget for work to be done before budgeting fo
r results to be achieved.



Enforce formal contracts in the market sector before introducing performance contracts in the public
sector.



Have effective financial auditing before moving to performance auditing.



Adopt and implement predictable budgets before i
nsisting that managers efficiently use the resources
entrusted to them.

Source: Public Expenditure Management Handbook. IBRD. 1998.

7

success.
6

Second, it provides a more structured approach to the planning and implementation of
reforms that wo
uld help inform and coordinate the reform efforts of governments, IFIs and
donors,
i.e.,
assist greatly in change management
.
7

Third, by sequencing into the medium term,
the approach provides a

vision


of the end result
8

that, while it cannot be attained
in the short
-
run, can be achieved by well identified smaller steps, again assisting in

selling


and sustaining
the reform effort. Fourth, is the practical argument: moving step by step, and completing one step
before beginning another, is more realistic i
n developing countries which are usually severely
capacity and resource constrained.

Although increasingly popular there have been some detractors of the platform strategy.

Some

within

IFIs, like

the World Bank
,

have found the above

arguments compelling, a
nd have
adopted the platform approach in support of large high profile PFM reform programs, such as
those in Kenya and Cambodia.
9

However, there have been a number of criticisms, aimed both at
the approach itself and how the approach has been applied.
10

Fir
st, some doubt the validity of
technical arguments for identifying some processes as more important, on which other PFM
processes depend, for inclusion in successive platforms.
11

Second, given each platform is an
amalgam of various reform activities it has
been argued that it is often difficult to agree with the
definition and objective of each platform, or hence reac
h consensus on its achievement.
12

Third,
the approach is often criticized as being overloaded in reform actions and over
-
engineered in
design, h
ence difficult to manage, and therefore prone to failure and reform fatigue
.
It should be
noted that as the approach has become more popular, the platform approach has also attracted
criticisms that could be applied to almost all other reform strategies, and the term itself has
perhaps been indiscriminately used as a way of packagi
ng reforms to better

sell


them to donors
and country authorities.

There have been
attempts to accommodate other approach
e
s into the platform strategy.
Tommasi (2009), while broadly endorsing the platform approach, has made the case that most
attention s
hould be paid to the first platform of the reform process in terms of planning and
implementation. For him, this first platform is the most relevant for most developing countries



6

For example, it is difficult to develop an adequate external audit function, without an adequate accounting system
in place. As such, the technical argument could be viewed as a variant of Schick’s

basics first


approach.

7
As Taliercio puts it,

given li
mited capacity, reform is broken into pieces for easier digestion
.”

(Taliercio, 2010
p.19).

8

What Schick calls

joining the dots
.”

9

For a review and critique of both these programs, see Allen 2009.

10

Note the warning by Schiavo
-
Campo,

Certainly, we need

to think in terms of complementary packages of
budgeting improvements

and to that extent the platform analogy is healthy and useful. But it is only an analogy,
and if pushed too far into prescription it risks becoming dysfunctional…
.” (
Schiavo
-
Campo, 2010
, p.7
)

11

This is equivalent to the controversy over what const
itutes “basics” in the “
basics

first”

approach.

12

As Richard Allen, so eloquently puts it:

If it is not possible to agree on a rigorous definition of the goals of the
reform strategy that are b
eing pursued

and a unique measuring rod for evaluating whether or not these goals have
been achieved

then the whole concept of platforms becomes

operationally meaningless, and collapses, much as a
building or bridge would collapse if the architects and con
tractors could neither agree

on how the foundations

of
that structure should be built, nor on a way of measuring when such work had been completed

to the required
specifications.”

(Allen, 2009, p. 18)
.


8

and should aim to ensure the

basics


are in place
.
13

While admitting there is

no agreed
definition of the

basics
,


he attempts for each sub
-
sector of the PFM system to separate

basic


from

beyond basic


priorities. To accomplish this he uses the PEFA analytical framework and
associated indicators. For each of the PEFA indicators

he delineates the desired rating, or target
score, that would suffice to ensure a

basic


level is attained in the PFM process covered by that
indicator. Consequently he takes the definition of basic PFM processes considerably further than
that attempted
by Schick. At the same time, he is less rigid from a sequencing viewpoint. The
first

basics first


platform he sees as being dependent on the country context in a number of
ways: since in all likelihood not all deficient basic areas can be tackled, the ch
oice of basic
reform priorities will be country specific
;
14

interconnections between processes may depend on
the country’s institutional context so priorities cannot be solely technically determined
,
15

and, in
terms of practicality, it may be necessary to incorporate on
-
going reforms that go beyond the
basics.

There have also been other approaches to using

the PEFA framework to assist in
sequencing decisions.

Another recent attempt to use the PEFA frame
work to assist in the
sequencing of PFM reforms has been advanced by Quist (Quist, 2009). He adopts a broad
platform approach founded on the stages in PFM system development, which he identifies with
various PFM objectives: aggregate fiscal discipline
;
16

ef
ficient service delivery; and strategic
allocation of resources. It is notable that his first platform delivering

fiscal discipline


is very
close to what most previously discussed contributions would subsume in the

basics first
approach

.
17

He then uses
PEFA indicators to define these platform capabilities
.
The novel part
of his approach, he calls the

PEFA Reform Sequence Model Framework

, is to technically
define the sequence of reform activities by looking at the inter
-
dependencies in the various PFM
s
ubsystem elements to identify the impact of each element on overall PFM performance, after
taking into account the number of

broken


PFM functions (i.e.
,

those scoring D, or N in PEFA
rating) and the number of well functioning activities (generally A and
B rating). He uses the full
PEFA 74 sub indicators to determine interdependencies.

He reviews the backward linkages, to
define what are called

activity chains

, i.e.
,

the number of activities that will have to be carried



13

These basic measures are generally those

focusing
on aggregate fiscal discipline, due process (including
compliance), and minimum requirements for accountability and transparency


(Tommasi, 2009, p. 30).

14

Although he does insert his own priority preferences in terms of priorities: for example, satisfacto
ry budget
coverage, adequate administrative a
n
d economic budget classification, clear accounting procedures and clear
definition of the responsibilities in budgeting (Tommasi, 2009, p. 26)
.

15

He gives the example in developing countries where formal rules
are properly designed but where informal
arrangements and gaming with the rules are rampant. In this context there is no point in further amendments to the
existing rules. Rather the priority should be increased transparency and improved external accountab
ility
mechanisms.

16

It should be noted that objectives are defined in a systems
context based on a simple input
-
output framework:
input, generates activity that produces an output that produces an outcome, and the outputs and outcomes have
feedback loops d
etermining future inputs. Thus fiscal discipline us defined in theses terms: sustaining the function of
the system; improving input controls and predictability; building activity capacity; strengthening output controls;
achieving regular, timely and accura
te feedback (Quist, 2010).

17


The first step for reform must focus broadly on sustaining the function of the system. This translates into a focus
on improving input controls and predictability, building activity capacity and strengthening output controls,
and
achieving regular,

timely and accurate feedback”(Q
uist,2009, p.10).

9

out to address the requirements for implementing a reform for any one given PI. These are
regarded as the pre
-
requisites for reforming this PI, or

reform effort
,


that is measured by the
total activity chain less those PIs that are functioning well. He al
so reviews the forward linkages,
asking if a given PI were to be made fully functional how this would positively impact all other
PIs,
i.e.,
an indication of the PI
'
s importance to th
e PFM system as a whole, this he

calls

reform
impact
.


His sequencing st
rategy is to opt for

fast wins


by choosing the PIs with the shortest
activity chains first.
18

Note by employing a country’s PEFA ratings as the basis of the model, his
is a country specific approach to the determination of reform sequencing.

Attempts to e
mploy
a purely technical sol
ution to sequencing problems have been
questioned.

It should be admitted that Quist'
s apparently technical approach to sequencing has
met with some criticism.
19

The use of the

system approach


in this model framework often
appears forced, but is necessary in order to define dependency within different PFM elements,
and certainly the description of the model’s derivation of activity chains and determination of the
interdependencies could be disp
uted. Some purists would view a system as denoting that every
element is dependent on the other, hence the idea of giving priority to one can be viewed as a
distortion of the system model. Similarly, the use of PEFA indicators is not without problems.
All
74 sub
-
indicators are treated equally, although some can be considered derivative indicators
measuring the results of how other indicators are performing.
20

Moreover, since some ratings
span different platforms they require some subjective allocation to dif
ferent platforms
.
This is
accommodated by his division of countries into two levels of PFM development: level one,
operating traditional budgets devoid of program orientation; and level two, operating what he
calls

a policy based


budget system. This lead
s him to treat the PEFA indicators differently in
his model for the two groups, with at least some additional complications for its application.
21


However,
while

using technical criteria

provided by PEFA
,
Quist recognizes

the
importance of non
-
tec
hnical fa
ctors when deciding
sequencing.

Despite the initial impression
of Quist’s model as being rather mechanistic and overly focused on technical aspects of PFM
reform in its approach to sequencing, this is far from the case. While attempting to identify the
mos
t important PFM elements for reform by the importance of dependencies between different
PFM elements, this analysis of PEFA indicators is not the sole (and perhaps not the most
important) determiner of reform sequence. He explicitly introduces the time req
uired to



18

To assist in this sequencing exercise he constructs a sequence rating indicator that measures the priority of the
PFM element due to its reform impact on the whole system.
The sequen
ce indicator is inversely proportional to
“reform effort” and proportional to “reform impact” (
See Quist, 2009, p. 28
)
.

19

See, for example, Schiavo
-
Campo, 2010, p.5.

20

See Tommasi, 2009,

Annex II.

21

Moving from one budget system to another (
i.e.,
level
one to level tw
o) involves a structural change, what he
refers to as a “foundational” reform rather than an “ordinary” PFM reform..

As he puts it,

when

a country decides
to transition from traditional budgeting to policy based budgeting it should not use
its previous PEFA scores to
determine the appropriate sequencing, but may be better served by adopting the base case to represent a degree of
restructuring its entire PFM systems to accommodate the structural change
” (
Quist, 2009, p.18
)
. This begs the
ques
tion of what happens when the reform program is aimed at changing
only
some elements of the budget system
from one level to another.

10

implement a reform as an important factor in sequencing

this he designates as

the roll out
effort

. This is viewed as dependent on the institutions responsible for carrying out the PFM
activities, recognizing some activities are centralized, so roll out is limited, but other activities
require many participants in the PFM system, and hence roll out eff
ort becomes a critical
consideration in sequencing. Moreover, Quist is very clear that sequencing should

like all
change management, take into account a number of considerations: political as well as functional
technical factors, and formal as well as inf
ormal institutional elements, capacity as well as
governance structures


(Quist, 2009, p.7). Given the importance Quist assigns to these factors in
his model, these would appear to outweigh the purely technical determination of PFM reform
design. Indeed, Q
uist is quite explicit that his technical analysis should inform rather than
determine reform sequence.

The

emphasis

on
the PEFA framework
for
sequencing

decisions

is also apparent in the
work of Andrews.

Another recent study that reviews reform design pri
marily through the prism
of PEFA indicators is Andrew’s research on the effectiveness of PFM reforms in 31 African
countries (Andrews, 2010). Although not the primary objective of the research, the study does
throw some light on sequencing PFM reform, admi
ttedly in this sub
-
group of countries. Using
their PEFA scores, on all PEFA sub
-
indicators, he divides countries into five

leagues

, (1

5),
determined by their level of PFM development, from low to high. It is possible to interpret these
empirically deriv
ed leagues as identifying

platforms


for PFM reform. He then asks: what
would it take for low league countries to reach the scores of high league countries? He divides
the PEFA dimensions into three: those in which the scores would be easiest to reach, (
i
.e.,
the
nearest), he calls

low lying fruit

; those in which the scores are furthest apart, and hence would
be the hardest to attain, he calls

high hanging fruit

; and those in the middle range,

middle
lying fruit
.


In this way his analysis reaches a possible prioritization in the various PFM areas,
identified by the PEFA indicators. Moreover, because most African countries are scoring in the

basics


range of indicators, this does offer an empirical approach based on

country reform
experience, to decide which basic functions should come before others in reform design. It is
likely that the approach could be generalized, and if replicated in other areas of the world could
also yield some interesting insights into seque
ncing priorities.

T
here are some criticisms of this empirical approach.

T
here are a number of dangers in
reading too much into the sequencing derived from this study. First and foremost, the order of
reform activities is determined on the

ease


of implem
entation, not the impact of the reform
action on the overall PFM system. The choice of reform actions may thus be biased towards
reforms that require less discomfort to stakeholders or involve a few central stakeholders for
success

criticisms that Andrews
has leveled at previous reform approaches.
22

Also, how
countries are selected for the sample is important. His sample was determined geographically,
but despite this he noted the importance of the considerable variation in the context in which
reform took
place. Andrews found significantly statistical differences in PEFA scores related to
such external factors as:

growth rates of their economies; social, political and economic stability;



22

Discussed further in Section 7.

11

the share of non mineral income sources; the length of the period of b
road reform commitment;
and perhaps also, colonial heritage (e.g.
,

whether Anglophone or Francophone budget systems
were operated). He is aware that statistical association does not imply causality, or if there is a
causal connection the direction of that
causality is often uncertain.
23

However, perhaps he does
not stress enough that reform may not be unidirectional, i.e.
,

experience has shown that progress
can be easily reversed in developing countries.

However,
once again
the importance of non
-
technical
factors

is stressed
.
It should be noted
that, like other contributors reviewed previously, Andrews stresses the importance of country
context and non
-
technical factors for success in PFM reform. Given this different inter
-
country
environment, would it be l
egitimate or desirable that countries follow the same reform path?

Andrews, himself, sees dangers in doing so. He is quite critical of the fact

that the reform paths
of these countries has been very much determined externally by the donors, through what he
calls international

reform products

, based on a same
-
size
-
fits
-
all strategy. Moreover, he is also
critical of the content of these products a
nd the longer run viability of the strategies being
pursued. He notes, for example, that PFM reform has been distinctly more successful in the
earlier stages of the budget cycle (upstream stages), rather than the budget execution stages
(downstream). This
is reflected universally in African PEFA scores, regardless of a country’s
league. This he sees as a by
-
product of these international reform products that have focused on
changes in laws, regulations and processes, (

de jure


changes) rather than on chang
ing behavior
(

de facto


changes). In doing so, reform efforts have focused on a few central agencies like

the
budget department, treasury, revenue regulatory agencies, and central procurement agencies, etc.
This he labels as involving

concentrated


PFM p
rocesses. The international reform products
have tended to avoid those reforms that require changes to behavior that involve a large number
of budget process participants (
i.e.,
the

de
-
concentrated


processes). It is the latter processes,
generally found
downstream in the budget, he argues, that urgently require improving but also
tend to be the most difficult to do so. In this conclusion he mirrors the importance that Quist
assigns to the rollout effort discussed above.

WHAT
DOES THE LITERATURE
TELL US
?

Despite the seemingly different approaches adopted, and the undoubted controversy that exists,
there are some general themes that emerge in the literatur
e:

Generally, the importance of sequencing in the design of PFM reforms is endorsed, but in
different w
ays and to different degrees.

Almost universally

experts in the field argue against
the

big

bang


approach, rather they view

an incremental or step
-
wise approach

or the
sequenced approach

as more

viable
.
This idea of the importance of the order in which reform
activities are undertaken is the main thrust of the

basics first approach
.


It implies the first
reform steps should aim to set up basic financial discipline through processes associated with a
trad
itional budgeting model. In this way, the

basics first


argument tries to define what the first
step in the reform sequence should be. However, to some the approach has often been viewed as



23

These problems are magnified in cross country rather than time series studies.

12

rather negative, since the associated argument is that most LICs
have a long way to go before
they can even ensure the

basics


are in place, and hence by implication should shelve more
modern reforms such as program and results
-
oriented budgeting techniques. The idea of
sequencing reforms to ensure their success is per
haps most strongly stated in the so
-
called

platform approach
,


where improved integrated PFM processes can, once established, act as a
solid basis for building the reforms of the next platform. While this has the advantage of taking
countries beyond the b
asics, some see this as potentially dangerous, hence, critics such as Allen,
question this sequenced approach as being too a
mbitious and doomed to failure.
24

However, it is
also worth noting that there is nothing inherently contradictory in the platform and


basics first


approaches, if the first platform aims to establish

basic


PFM processes.

There are some hints at an underlying contradiction in prioritizing reforms.

I
n much of the
above discussion there seems to be an underlying contradiction in adopting the idea of
sequencing PFM reforms: namely, is there an underlying tension between viewing PFM as a
system, where all sub
-
systems are interlinked, and then choosing a
sequenced approach where
inevitably some subsystems are prioritized in the reform design? As indicated, some, such as
Quist, try to relieve the tension by designating some subsystems more critical than others
through an analysis of their linkages in the PF
M system. In this way he attempts to define a
country
-
specific sequenced approach to reform based on PEFA indicators. However, as
indicated, some remain skeptical of this approach, not the least because the characterization of a

system


that underlies the

concept of linkages is somewhat artificial, his heavy reliance on
PEFA indicators that are not designed for this purpose, and the inevitable qualifications that have
to be introduced into what some see as an overly technical approach to sequencing.

While
most experts in the field recognize the value of a sequenced approach, at the same
time most would also agree to disagree on the desired order of sequenced reform actions.

The lack of harmony in the literature has been noted above. For example, the questio
n as to what
should the first platform encompass has been answered in different ways in different countries.
This lack of consensus is most evident in the various interpretations on what constitute PFM

basics
.


Some define

basics


negatively as what they

are not (see Schiavo
-
Campo, 2010,
p.9)
.
25

Some have been bolder. For example, Schick’s initial paper laid out his view of

basics
,


which appear to be in line with traditional budgeting fiscal discipline. But even this checklist,
which seems innocuous enou
gh, is not without its detractors. As noted, Tommasi has gone
further, and in some detail, examining each PFM sub
-
system he has tried to define what would
be

basic


PFM processes and what would be

beyond basics
.


However, in many ways this
important exer
cise highlights the depth of the problem faced in reform design: even if we can
define the first step in a sequenced reform as

basic
,


we would not be able to tackle all basic
sub
-
systems at once, so how to prioritize between them? Are some basic processe
s more



24

While Allen has been critical, it should be noted at the same time as eschewing the sequenced approach at the

macro


system
reform level, he would see value in sequencing at the lower, sub
-
system level and identifies various areas where this has proved
su
ccessful.

25

Although, to be fair, he does offer a basic checklist of first priority PFM objectives: protect public money;
complete budget coverage; offer a multi
-
year perspective; improve public investment preparation and programming;
and provide some atte
ntion to results. (Schiavo
-
Campo, 2010,

p
.
12)
.

13

important than others? But if they are, (as Quist would argue), surely for the PFM system’s
viability as a whole there needs to be a balanced approach in reform. If one subsystem is
advanced much further than the others is this reform likely to be a
s effective without the support
of other subsystems? This, after all, is the logic of the platform approach.

It should be noted that most experts in the field are willing to accept the PEFA framework
as a method of assessing effectiveness in PFM reform pro
grams.

Some, (like Tommasi and
Quist), have turned to the PEFA indicators to identify which indicators best reflect what the

basics


should be and to use the ratings of indicators to prioritize reform efforts. Unfortunately,
PEFA indicators on many dimensions are graded between basic (D, C ratings) and best practices
(B, A ratings), but the majority could be characterized as describing

basic traditional budget
control processes (even with an A rating). However, even if agreeing on the use of PEFA, there
is disagreement between experts on required indicator ratings to reach

basic


standards
.
Another disadvantage as an aid to sequencing
decisions is that the PEFA framework does not
prioritize between its indicators, even though some are derivative (or what Tommasi calls

resulting indicators

).
26

However, perhaps the problem lies deeper than PEFA’s limitations.
Underlying the failure to ag
ree a universal order for sequencing reforms is a generally held
opinion that this would in any case be undesirable. This opinion is founded on two consistent and
interrelated themes that emerge in the literature reviewed: first, reform design should be co
untry
specific, and there are dangers in generalizing across countries; secondly, successful reform
design involves more than technical aspects of PFM, and other external factors are often critical
to success.

There seems a general consensus that there can

be no universal reform sequence, since
reforms should be geared to specific country circumstances.

This conclusion
could be
interpreted as

nullify
ing

any attempt

at generalization about sequencing.
27

However, if one
examines the literature, especially that

dealing with country case studies, it is not difficult
to
discern some agreement

on what should come first. The agreement is

admittedly

incomplete, as
the previous discussion on what c
onstitutes

basics


makes clear. N
one the less
,

the first steps in
sequ
encing reforms in many different countries in different parts of the world are not so widely
different as one might at first expect. Previously it was noted that Schick’s view of the first steps
in sequencing should be directed to what he would call fiscal

discipline. An examination of the
content of the first platforms in reform programs most usually quoted, that of Kenya and
Cambodia, reveals these first platforms generally also contain reform activities directed to those
areas that Schick would most prob
ably recognize as falling into his

basics first


category.
Similarly, the

evolutionary approach


favored in Ethiopia, contains as its first stage the



26

It has been noted that Quist attempts some prioritization, dividing countries into Level 1, (basic traditional
budgeting) and level 2 (policy
-
based budgeting); and Andrews identifies empirically priorities
for lower level
African countries to progress their PEFA scores to those of higher level ones.

27

Note

Schiavo
-
Campo’s warning that “
we must recognize that any prioritization must be country
-
specific and
reality based

,

and that

attempts at identifying a u
niversal sequence of actions are not only unrealistic, but also
undesirable as they carry the twin assumptions that the sequence is both linear and unidirectional
.” (

Schiavo
-
Campo, 2010,

p

4
)
.

14

establishment of basic traditional budget controls, not out of line with Schick’s view of basics.
28

So wh
ile in the sequencing literature there is much debate on the order of reform action, when it
comes to reforms in practice the differences do not appear so great. This suggests
one should not
stress too much the uniqueness of countries as undermining the id
ea of some logical
order in
sequencing

reform
.

What is evident is that although their first platforms usually contain activities
aimed at establishing

basic


PFM processes, countries differ in where they give priority,
favoring some basic PFM processes over others. The literature suggests these different priorities
perhaps can best be explained by non PFM factors
.

Almost all experts in the field stress non
-
tec
hnical factors as heavily influencing the design
and the success of reform initiatives in the PFM field.

Unfortunately, while there is general
agreement on the importance of non
-
PFM factors, there is no consensus on what they are, except
perhaps again that

they vary between countries in importance. The scope of the factors identified
in the literature is very wide. Allen (2008), complains that PFM reform advice is often poor
because advisors are unable to see beyond budgetary institutions and PFM systems. H
e argues
that it is important to recognize that

reforming institutions is a necessary but insufficient
condition for improvement; other important considerations are people, skills, organization, and
information


(Allen, 2009,

p.23). This echoes the argume
nts advanced by Schiavo
-
Campo and
McFerson, 2008. Brooke, in advancing the platform approach, highlighted the importance of the
political context, as well as four other important factors: capacity development; motivational
development; process development;

and institutional development.
29

For Tommasi the country
context includes,

among other elements, human resources capacity, current strengths and
weaknesses of the budget system, the administrative culture and the institutional and political
context.


(200
9,

p.

10). This agrees with Quist, (2009, p 7), who sees

the political context, the
maturity and level of activity by civil society, the level of donor harmonization, the degree of
ownership of reform strategy along with the political will, all factor int
o the sequencing of
reforms

30

As noted in the African context Andrews empirically associates a number of external
factors (growth, stability, revenue structure, length of reform commitment, colonial heritage)
with higher PEFA ratings. Given this common pos
ition, it is difficult to avoid the conclusion that
in reform design one may have to accept non
-
technical factors will trump purely PFM technical



28

The sequence starts with building a

transactions platform


(budget system, accounting, disbursement systems
and their supporting systems) to ens
ure efficient financial control
(Peterson, 2007).

29
Capacity development
denotes that people, institutions and societies are enabled to perform functions, solve
problems, s
et and achieve objectives.
Motivational development

considers how people can be carried through
changes.
Process development

addresses the aspect of organizational structure and how organizational change can
occur.
Institutional development

looks at aspect
s such as laws or policies, and also deals with institutional culture
that regulates behavior. It should be noted that Bergmann and Bietenharder when they surveyed experts in the field
of public sector accounting reforms found that there was a general acce
ptance that these “soft factors” were central
for sustainable PFM reforms (Bergmann and Bietenharder, 2010, p.6).

30

Moreover, he would also add,

internal political factors pertaining to appropriate incentives for the professional
cadres responsible for u
ndertaking the PFM reform activities that must be considered. These too are always complex
and require a subtle appreciate on the roles, inter
-
relationships and power levels of the reform leadership


P. 7
.


15

factors in determining priorities and the sequence of reform actions, both the order in which they
are underta
ken and the timing required for their completion.
31


There appears a general acceptance of
the
view that success and failure in PFM reform
should be judged by the outcome of reform action in achieving PFM objectives.

In the
design of reform programs, nearly all experts stress the importance of setting milestones of
success by indicators of outcomes to be expected by PFM reforms. This focuses on the

deliverables


of the PFM system. Usually care is taken to specify what

is to be the expected
outcome of meeting

basic


PFM processes, or the expected outcome of achieving platform
targets
.
This appears to be a common role assigned by most experts to the PEFA process
--
an
assessment tool founded on a framework defining the de
sired outputs/outcomes of a well
functioning PFM system. It is also recognized in the literature that not all desired PFM outcomes
are relevant for different countries at different stages of the development of their PFM systems.
Most contributors recognize

that at different stages of development, countries operate what are
essentially different types of budget system: moving from a traditional budget environment
concerned with the control of inputs to

policy based budgeting


more concerned with outputs
and

outcomes. Thus Schick matches his

basic


PFM functions to different types of control
required by a system that centralizes budget management, concentrates on inputs (individual
items of expenditure) and is more concerned with compliance with line budgets
, and civil service
rules and financial regulations.

For Tommasi the basic PFM system focuses on fiscal discipline
ensured by control over fiscal aggregates, emphasis on due process, compliance and probity,
with basic legislative scrutiny and accountabilit
y and hopefully with an adequate degree of
transparency (Tommasi, 2009, p.22). As noted, Quist divides countries into two levels, where the
first level operates a traditional budget system that would not be markedly inconsistent with that
described by Tomm
asi.
32






31

This Dorotinsky would argue is necessary, since reforms need to be feasible in the given political context, and it is
i
mportant that they be defined

by the

patient


what is the most desir
ed treatment that this required

determined
,

he suggests
,

by what is

the weakest link in the PFM subsystem (Dorotinsky, 2010)
.

32


The first platform (or series of platforms), consistent with the achievement of fiscal discipline,

will therefore
include such reform activities so as to address budget comprehensiveness; predic
tability of revenue receipts, budget
releases, procurement and expenditure steps; control
-

with

input controls being addressed prior to output controls;
transparency; accountability; and the frequency, timeliness and accuracy of reconciliation procedures,

financial
reporting and accounting; and finally external audit and oversight


(
Quist, 2010, p.18
)
.

16

CHAPTER II:
A

FRAMEWORK
FOR INTERPRETING

THE SEQUENCING LITER
ATURE


Summary:
This chapter revisits the meaning of sequencing, and suggests an analytical
framework based on two main principles. First, that two fundamental dimensions of sequencing
should be considered together: the order of reform actions and their timing. It is argued the
order of reform actions can follow an historical perspective but timing, or the pace of reforms,
will depend more on non
-
technical country specific conditions
.
Se
condly, regarding

the

order of
reform activities a hierarchy in sequencing decisions is proposed: the higher level priorities
should be agreed before going on to the lower order
,

more technical
,

aspects of sequencing
individual reform actions. Using this f
ramework the literature on PFM sequencing is
reinterpreted. This identifies some areas requiring further clarification, and indicates how they
will be addressed in this
Background P
aper
.

A SUGGESTED FRAMEWOR
K FOR
REVIEWING THE SEQUEN
CING LITERATUR
E

There is the need to impose some structure on this wide ranging literature.
In attempting to
distill guidelines from this seemingly disparate and wide ranging literature, it was felt necessary
to offer a conceptual frame
work in which to fit the various arguments encountered. This
framework rests on two main premises. First, sequencing reform involves more than just
deciding the technical order in which reform activities should take place. Second, sequencing
reform activit
ies cannot be divorced from the ultimate objectives of reform, so without some
agreement on the latter it is difficult to agree on the former. An attempt will be made in the
following section to reinterpret the literature within this conceptual framework.
It is hoped that in
this way to stress the surprising degree of consensus to be found in the literature, as well as
highlighting the main areas of discord, and pointing to further issues that should be addressed
.

The definition of sequencing should recogn
ize two d
imensions
.
It is suggested that as a first
step it may be useful to revisit how

sequencing


has been interpreted in the literature. While
there are special interpretations of this concept in science, music, and mathematics, a most
general definition found in the dictionary for

sequence


is as a chain of events. So when talking
of sequencing PFM

reform it would appear we are trying to define a chain of PFM reform
actions
.
Within that definition it is also possible to identify two associated dimensions: the
order

in which actions are taken and the
timing

of the actions. Indeed, the sequencing lite
rature is
peppered with references on the need to be realistic about the time taken to complete reform
activities, and factors that may affect this timing. This is often reflected in discussion of the

pace


of reform

(see Schiavo
-
Campo and Tommasi,

1998)
.
33

For example, Browne also
emphasizes pace, stressing that


phased implementation

of any sequenced action is important


in ensuring that reforms will fit the absorptive capacity of the system and not cause reform



33

They stress that

strategic attention should focus on identifying areas where it is feasible to move very fast, and
areas where it is essential to build s
lowly and carefully the solid institutional foundation required for sustainable
reform
.”

(Schiavo
-
Campo and Tommasi, 1998, p.109 )
.

17

fatigue


(Browne,2010, p. 12). She recom
mends that in the design of a sequenced reform there
should be periodic digestion and consolidation periods built in to ensure changes are internalized.
When discussing the pace of reform DFID sees this as heavily influenced by whether reforms are
driven b
y crisis, as well as if they are internally and externally driven (DFID, 2001, section 9.1).
In the literature it is also possible identify arguments that the time taken to implement reform
activities should influence the order in which they are undertaken
, for example, in emphasizing
the need to consider roll
-
out effort in sequencing priorities also suggests reform actions that will
have priority (Quist,

2010)
.
It will be argued that these two dimensions

order and timing

should be accommodated in any attem
pt to sequence PFM reforms, and that decisions on both
dimensions will need to be consistent to assure success. Previous discussion on reform
sequencing often has failed to explicitly differentiate these two dimensions, and has perhaps
tended to concentrat
e more on the order of reform actions.

T
he
ordering
of reform actions should be thought of as a two
-
stage process.
First, at a higher
level of decision making, it consists of getting the overall reform priorities correctly ordered.
These reform priorities should be decided by what the PFM system is designed to deliver (or
w
hat is the outcome to be achieved by the PFM s
ystem as a whole
?
). Secondly, once the high
level reform priority is decided, at a lower more technical level of decision making, sequence
consists of ordering reform actions to achieve this priority or deliverable. It is at this lower level
that detailed
sequencing of reform activities is most justified. It will also be argued that if the
higher level prioritization is wrong, then sequencing reform actions at the lower level is unlikely
to be successful.

T
he second dimensi
on,

the
timing

(or pace)

of refor
m actions is

often acknowledged as

critical to successful reforms.

Getting the timing wrong, both of when to commence and how
long to allow for a reform action, will undermine any successful sequencing of reform. Further, it
should be recognized that the t
iming of reform actions often depends critically on non
-
technical
factors that are external to the PFM system, and as such will inevitably be country specific. This
has two implications. First, while it may be possible to generalize across countries about
the
desirable order in which PFM reforms should be undertaken, since timing will be country
specific, it is not likely to be possible to generalize across countries about reform sequencing
overall
.
Second, sequencing cannot be considered solely a PFM techn
ical matter outside each
country’s political economy of reform. It is difficult to avoid the conclusion, therefore, that to
come to implementable decisions on sequencing requires these external non
-
technical issues be
addressed directly.

A hierarchical app
roach to s
equencing

is proposed.
E
xploring a

two stage approach to
ordering reform actions, there seems some
broad
international agreement about what

the

top
level priorities

of a PFM system should be. This agreement is based on what the PFM system is
designed to deliver.
There are many variants on the theme, but
it is argued here that in

an

"
ideal
"

PFM system there are at least three main

deliverables

.
First, there should be

basic financial
compliance or fiscal

control

enforced by

the system.

Second, there should be
processes in place
to
ensure aggregate fiscal discipline
, in the sense that the PFM system can
support

18

macroeconomic stability in the economy
over the business cy
cle
, as well as longer
-
run macro
fiscal sustainability.

Third,

given the previous two deliverables,

the
PFM
system should be able
to
attain agreed sector policy priorities in an efficient and effective way
.

It is evident that the view of the scope of PFM adopted in this paper, and in particular the
identification of its deliverables, is at variance with some other accepted interpretations.
Specifically, over the years a particular perspective on the objective
s of PFM has emerged and
become part of the landscape, often used but not much questioned. This perspective focuses on
the threefold pillars of a good budget system that was introduced in the 1999 World Bank PEM
Handbook, namely: aggregate fiscal disciplin
e, strategic allocation of resources, and efficiency in
resource use. While not denying the threefold classification has its uses, it can be argued that it is
far from sufficient in discussing the deliverables of a PFM system or for modeling PFM reform.
Th
e main deficiencies are twofold: it emphasizes policy objectives not management objectives;
and is too narrow in scope, concentrating on budgeting but not the whole PFM system.

It can be agreed that t
he budget formulation process

should focus

on resource

allocation
decisions determining government policy at three levels
, namely:



Top level decisions determine how resources are to be allocated between the government
and private sector, revolving around the desired size of government. This determines the
ove
rall resource envelop available to the budget. If functioning properly this aggregate
use of resources by the government sector will be compatible with overall
macroeconomic balance in the economy and any resulting debt levels will be sustainable
over the
longer
-
term.



Middle level decisions determine how this overall envelop is to be divided between the
different sectors/ministries to meet policy priorities. If functioning properly the budget
process will ensure that resources are moved to high priority are
as away from low
priority sectors and programs.



Lower level decisions determine how resources will be divided within the sector/ministry
envelop to best meet policy objectives. If functioning properly, resource allocation
decisions at this level will be ma
de so that policy objectives are reached in the most
efficient and effective way.

All these resource decisions are usually reflected in the budget document, which represents
perhaps a government’s most important policy document.

However, some qualificatio
ns are
in order. It should be noted that often the budget is an incomplete policy document since
government policy can be pursued by other means: extra
-
budgetary funds, quasi fiscal
operations, regulations, etc., all such activities should be included from

a
public financial

management

perspective. Moreover, few countries actually display such a pristine strategic
budget process envisaged above. The three levels of decision making tend to be taken together
not separately. Bottom
-
up demands often determine o
verall government size; how sector
envelopes are decided often determine how ministries allocate their resources, etc.

Seldom in
19

practice is the budget formulation process broken down into these three strategic decision levels,
but they are intertwined in some negotiated end result that is politically acceptable.

However, the most important qualification to adopting this approach t
o modeling PFM
reform is that budgeting (preparation, presentation, approval) is only one component of
PFM.

The PFM system

represents all the regulations and procedures that ensure that the
threefold policy decisions previously described are fully complied

with. That is, apart from
ensuring the budget preparation process is effectively functioning as described previously, there
must be in place a system to execute the budget that ensures resource allocation decisions are
complied with, monitored and reporte
d, and treasury operations that ensure this execution takes
place smoothly in terms of cash availability, and is financed efficiently through debt
management operations. Adopting this wider perspective it is possible to identify the

t
hree main
management o
bjectives or “deliverables” of a PFM system used in this background paper:




First, the system should enforce basic financial compliance.

Without this basic feature it
is impossible to deliver gover
nment spending in line with resource allocation decisions
at all
three levels that have

been politically agreed, and so use the budget as a tool of government
policy.



Second, there should be processes in place to ensure aggregate fiscal discipline, in the
sense that the PFM system can anticipate and adopt any c
hanges required to fiscal
aggregates to counter macroeconomic imbalances and help ensure overall stability in
the economy.

This requires budget planning to ensure macroeconomic stability

not only
from the perspective of

the current fiscal
year

but over the

business cycle, and even for

longer periods to ensure this stability can be sustained. From the

poli
cy viewpoint
, basic
macro fiscal stability is often considered a precondition
and safeguard to successfully
delivering other government policies reflected
in budget allocation decisions.




Third, given the previous two deliverables, the PFM system should put in place
procedures to be able to attain agreed sector policy priorities in an efficient and
effective way.
This usually requires the adoption of a progr
am oriented approach to the
budget, focusing on the outputs rather than the inputs of the budget system, and the use of
performance information and processes to enhance program managers' delivery of desired
policy objectives.

From this is derived an overal
l picture of a more advanced PFM system shown in Figure 2.1,
describing the management processes required to deliver strategic policy decisions at all three
levels. The three fold deliverables of this system underlie

many of the currently used
internationa
lly accepted tools, such as PEFA, used to assess how successfully a country’s PFM
system functions.

Indeed it should be noted that the majority of PEFA indicators measure the
degree of financial compliance, an objective completely ignored in the convention
al threefold
classification of objectives.




20

Figure 2.1

Schematic Overview of the PFM Conceptual Framework

based on

Three Main Deliverables


Aggregate
resource envelop
decided
Strategic
allocative
decisions (resolve
conflicts)
Strategic
Sector plans
(defines resource
needs)
1. Financial compliance/fiscal control
Rests on adequate: regulatory framework; coverage of govt.
operations; transparency based on good accounting; supporting IT
A
ssures fiscal
aggregates are
achieved as planned
A
ssures sector
allocations used for
correct purpose
2. Medium
-
term
planning/ Sustainability
analysis
3. Program &
performance budgeting
Achieves
macroeconomic
stabilization
Achieves
Efficiency &
effectiveness
Policy
decisions
PFM Requirements to deliver policies


THE HIERARCHY AMONG
TOP LEVEL PFM PRIORI
TIES

The history of how PFM systems have actually developed also suggests a hierarchy among
these top level
deliverables
.

If we examine how the advanced PFM countries developed their
PFM systems, we find, (accepting particular country idiosyncrasies mentioned b
y Allen)
,
34

a
similar progression as that described above. First, countries pursued
financial compliance
, in this
way to ensure control over revenues and expenditures so that they are consistent with annual
budget appropriations and are undertaken in
accordance with budget legislation usually
buttressed by detailed financial regulations.

Generally such PFM systems focused on detailed
line
-
item budgeting with the emphasis on ensuring the most economical use of inputs
.
For this
basic control to be effect
ive requires the budget to be as comprehensive as possible to include all
government spending, with centralized controls in a strong finance ministry, the operation of
centralized cash management (a treasury single account), a regular budget calendar, time
ly



34

It should be noted that Dorotin
sky agrees with Allen on the idiosyncratic development of PFM systems. Namely
for him PFM s
ystems

were developed gradually, incrementally, addressing one weakness at a time, always
responding to the needs of policy makers
.”

(Dorotinsky,

2010).

21

(usually cash) accounting, regular reports, strong internal controls, and an active external audit
function reporting to the legislature
.

Once this basis of

financial discipline was achieved, historically with the rising importance
of government activi
ty in the economy and with the spread of Keynesian ideas, more
emphasis was placed on controlling fiscal aggregates over time
.
PFM systems were
developed so that fiscal aggregates could be adjusted to attain aggregate fiscal discipline so as to
ensure macr
oeconomic stability and sustainability
. It was realized that both objectives involved
fiscal planning over the business cycle. This led of necessity to a break from rigid annual budget
planning, and to embrace a medium term perspective in fiscal planning.
In terms of PFM
processes this was associated with the development of improved macroeconomic and fiscal
forecasting, timely reports on fiscal aggregates, medium term fiscal and budget frameworks, and
debt sustainability analysis. For developing countries,
it has also encouraged better investment
planning, taking account of future recurrent cost implications of present capital spending, and in
better programming and managing of aid flows.

Latterly, more especially in OECD countries, PFM systems have been
adapted to ensure
better policy delivery, with an emphasis on outputs and outcomes of publicly provided
services.

This
emphasis on efficiency and effectiveness
in service delivery has resulted in PFM
emphasizing strategic planning, program budgeting, use o
f performance indicators to monitor
and evaluate, and more decentralized management systems, typically associated with a move
from cash to accrual basis accounting (see Chapter VIII)
.

This progression involved a change in emphasis and a widening of

PFM ob
jectives, not
replacing one with another
.
While countries have varied greatly in the universality in which
they adopted these three main PFM functions, (and even within OECD countries the process
continues), the history of reform in this area generally sho
ws a widening of PFM objectives in
the direction previously described. It also shows that at each stage of developing their PFM
systems the new objective, or deliverable, was added and did not replace other deliverables. That
is, the later emphasis on valu
e for money and policy delivery still rested on financial compliance
and aggregate fiscal discipline. This view of reform priorities based on this historical perspective
is illustrated schematically in Figure 2.
2
. The pyramidal form stresses that the
deliverables
should be regarded interdependent: financial compliance supports the other deliverables, both
macro stabilization and efficiency and effectiveness; and in turn macro stabilization supports
efficiency and effectiveness and eases the task of ens
uring financial compliance. From the PFM
management perspective, these stages represent the development of different skills and
processes: those required for traditional compliance oriented annual budgeting will have to be
supplemented by further skills to

deliver multi
-
year macro
-
stabilization and sustainability
objectives; and these in turn will have to be supplemented by other skills and processes to deliver
increased efficiency and effectiveness in service delivery.


22

Figure 2.2 Reform Priorities
Primary Objective/deliverable Type of budget system
3. Efficiency &
Effectiveness
2. Macro
-
fiscal stability
& sustainability
1. Financial Compliance
Concerned with outputs and
outcomes
Emphasis on results/policy
delivery
Decentralized management style
T
Traditional budget concerned with
emphasis on input control
Annual orientation
Centralized management style
Traditional input control,
with multi
-
year orientation
Emphasis on macro policy
Centralized management style

Historical precedence should be used as a guide in sequencing at the highest level.

If one
accepts this broad view of PFM development, one must ask why reinvent the sequence of budget
reform?

Would it not be safer to adopt th
e sequence suggested by historical precedence:

first,
create a PFM system that delivers financial compliance/ fiscal discipline; second, once achieved,
develop the PFM skills and processes so as to be able to determine and adjust fiscal aggregates to
ensur
e macroeconomic stability/sustainability; third, once this is achieved move to getting better
value for the money spent in terms of better achieving sector policy objectives and attaining
more efficiency in delivering services
.
These three stages in PFM de
velopment, it is suggested,
should guide the ordering of higher level priorities in shaping the sequencing of PFM reforms. At